In recent years, study about separation and purification of a specific gas from a mixed gas using polymer membranes as gas separation membranes has been actively conducted. For example, attempts have been made to produce oxygen-enriched air and to make use of it in fields of medical care, fuel systems, and the like. Then, the gas separation membranes used for these applications are required to be high in both gas permeability and gas selectivity to the specific gas. Further, depending on the usage environment, it is also required to be excellent in heat resistance, chemical resistance, strength, and the like.
Under such circumstances, various gas separation membranes have hitherto been proposed. For example, Patent Document 1 (JP-A-10-99664) proposes a carbonized membrane for gas separation that is obtained by molding a cardo type polymer having a predetermined structural unit into a predetermined separation membrane form to form a separation membrane precursor, and heating this separation membrane precursor in an anaerobic atmosphere to carbonize it.
Meanwhile, for polymer membranes, there is a trade-off relation between the separation factor of two kinds of gases and the permeability coefficient, and the presence of an upper bound region has been advocated (see Non-Patent Documents 1 and 2).
Further, the present inventors have previously proposed a gas separation membrane composed of an organic-inorganic polymer hybrid having a hyperbranched polyimide moiety and an inorganic oxide moiety, which are integrated by a covalent bond to form a composite structure (see Patent Document 2).    Non-Patent Document 1: Lloyd M. Robeson, “Correlation of separation factor versus permeability for polymeric membranes”, Journal of Membrane Science, the Netherlands, Elsevier Science Publishers B V, vol. 62 (1991) pp. 165-185    Non-Patent Document 2: Lloyd M. Robeson, “The upper bound revisited”, Journal of Membrane Science, the Netherlands, Elsevier Science Publishers B V, vol. 320 (2008) pp. 390-400